Method for a Dynamic Perpetual Encryption Cryptosystem
A dynamic computer communication security encryption method or system using an initial seed key and multiple random number generators of a specific design, whereby a sequence of independent random entropy values is produced by one set of random number generators and encrypted along with the message stream using the initial seed key, or the output of a second set of random number generators initialized with the initial seed key, and following the subsequent transmission of the variable encrypted entropy/message block, the entropy values are used to symmetrically or identically augment or increase the current uncertainty or entropy of the cryptosystem at both the sender and the receiver, prior to the next encryption block operation. The encryption process effectively entailing the use of multiple encryption ciphers, and the entropy augmentation process entailing the encryption or application of various logical mathematical operations on the already dynamic but deterministic internal state values of the second set of random number generators, effectively altering their deterministic outputs in a random probabilistic manner. Random length message value sequences from one or more data sources is combined with one or more random length entropy value sequences from an independent source, following which the entropy “updates” may also be used to alter, or change any cryptosystem variable, value or component in a randomly determined manner. In addition, whilst ensuring synchronization, the random entropy sequences also serve to “pollute” the cipher-stream and thereby hinder most current forms of cryptanalysis, whilst simultaneously injecting additional entropy into the cryptographic system and allowing for its propagation to affect any connected system nodes, and thereby introducing unpredictable entropy into the system pseudorandom number generator outputs, and thereby ensuring the perpetual generation of unpredictable random numbers. Super-encryption mechanics are independent of the user data, simple, fast and efficient, and can incorporate compression, error correction and asymmetric encryption authentication routines. But most importantly, super-encryption ensures resistance to brute force attacks (not possible to verify if a message was even sent), an ability to exceed “perfect secrecy” requirements, and an improvement on previous super-encipherment design, since overhead can be dramatically reduced from 100% overhead. Communication links previously established by system nodes with central authorities may be used for secure node authentication and registration, whilst allowing the central authority to broker and synchronize communication channels and providing mutual authentication and other security functions between the system nodes.
U.S. Pat. No. 1,310,719—Gilbert Vernam—“Secret Signaling System”
U.S. Pat. No. 5,412,730—Michael F. Jones—“Encrypted Data Transmission System Employing Means for Randomly Altering the Encryption Key”
U.S. Pat. No. 7,376,232,—Hamdy Soliman—“Computer System Security Via Dynamic Encryption”
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENTNot Applicable
BACKGROUND OF THE INVENTION(1) Field of Invention
The present invention relates to data transmission and computer security systems and more particularly, to multi-cypher encryption and decryption, multi-source random number generation and to systems for transmitting multiple encrypted data streams.
In particular, the present invention relates to a random number generation and encryption method or system which uses a distributed framework of multiple random or pseudo-random number generators producing infinite length multiple keystreams, to support a dynamic super-encryption cipher-stream (a continuous sequence of variable length ciphertext segments each encrypted with a unique key) which is used to transmit data between system nodes, and also transmit random entropy updates which are used to perpetually augment the entropy or uncertainty of the cryptosystem mechanics, variables, configuration, state or the composition of the random number generators used to generate the required random number outputs.
(2) Description of the Related Art
Cryptography provides a means of security for transmitted data by transforming messages from “plaintext” to “ciphertext” using a secret “key” and an encryption and decryption algorithm.
This invention relates specifically to “true” secrecy cryptosystems, in that the meaning of the messages being transmitted are concealed by a cipher or code and the adversary is assumed to have any special equipment necessary to intercept and record the transmitted signal; as defined in the document—Shannon, Claude (1949) “Communication Theory of Secrecy Systems”, Bell System Tech. J., Vol. 28, pp. 656-715.
In typical “symmetric” encryption usage scenarios, the transmitted data is encrypted by a sender using a unique key value (such as a 256-bit binary number) and unscrambled at the receiver, by a reverse process that requires the same key value to be known, whereas with “asymmetric encryption”, different key values are used which are mathematical inverse functions of the first unique key value, such as ¼ and 4/1. Since these keys are almost always shorter than the messages they encrypt, a means of applying the key to longer messages is used. With block cyphers, the key is in essence repeated with every block of message data that is encrypted (may include a transformation of the key between blocks), whereas with stream ciphers the key is applied to a pseudo-random number generator (hereinafter called “PRNG”) to produce a long keystream which is then used for encryption and decryption purposes.
Ordinarily, the strength of a cryptosystem may be considered to be directly proportional to the length of the initial key, but this assumption is subject to the condition that the ciphertext is shorter than a value that Claude Shannon called the “unicity distance”. Cryptographic algorithms which produce ciphertexts which exceed the unicity distance, which is the length of the initial key plus the minimum length which could be used to convey the message, are inherently insecure, and are guaranteed to produce a single correct decipherment. Since the encryption strength of any encryption algorithm decays in proportion to the length of the ciphertext which exceeds the “unicity distance”, statistical decryption techniques such as equivocation are far more efficient than brute force decryption.
A simple and secure cryptosystem which is not subject to equivocation attack risks, has been around since 1882. The “one-time” pad (hereinafter referred to as “OTP”) or Vernam Cipher, was originally invented by Frank Miller and patented in 1917 by Gibert Vernam (U.S. Pat. No. 1,310,719 entitled “Secret Signalling System”). The OTP has the distinction as being the only cryptosystem to date to have been mathematically proven to be “perfectly secret” (as defined by Claude Shannon) if used correctly, in that an adversary is unable to compromise the cipher given infinite time, infinite computing resources, and any number of plaintext/cyphertext pairs. It is the object of this invention to introduce a new “information theoretically secure” cryptosystem.
The OTP's “perfect secrecy” is dependent on the following conditions being met: (1) The key used must be truly random—the key must therefore be unpredictable or insignificant in every sense of value, position or distribution, (2) The key must be at least as long as the plaintext, if not longer, and never repeated in whole or in part, and (3) The keys must be kept completely secret during generation, distribution, usage and storage.
Claude Shannon, in dealing with secrecy cryptosystems in general, asserted that the OTP complied with the requirements for “perfect secrecy”—namely that the ciphertext can be translated into any plaintext of the same length, all of which are equally likely, if a truly random key was used. More specifically, he defined “perfect secrecy” as meeting the criterion that “the a priori probability of a plaintext message M is the same as the a posteriori probability of a plaintext message M given the corresponding ciphertext”. Further, he explained that perfect secrecy is possible but requires that the keys must be generated at the same or higher rate than the messages, and be at least as long as the messages. Strictly speaking, by the above definition, perfect secrecy can never be obtained using block cyphers or seed-fed stream ciphers, since the input entropy is limited to the seed or key size.
Despite these cryptographic strengths, practical problems have prevented OTPs from being widely applied in a practical manner, including: (1) The Random Number Generation Problem—The problem in generating unpredictable infinite length one-time pad sequences, which is a serious problem if one is using single independent pseudo-random number generators (hereinafter called “PRNGs”), (2) The Distribution Problem—in that the required OTP values must be securely generated and exchanged between the sender and receiver, and be at least as long as the messages to be sent, and (3) The Key Management Problem—in that the OTPs must be kept secret and be immediately and irrevocably disposed of in a secure manner.
In addition to the problems mentioned above, the OTP also suffers from the following issues: (1) known ciphertext length, (2) ciphertext readily visible for analysis, (3) lack of user authentication, (4) lack of message verification and error correction, (5) ciphertext position is relative to plaintext.
In addressing the Random Number Generation problem with regards to “perfect secrecy” systems and the “one-time pad”, any numbers that are generated and used as keys in a cryptosystem must be truly random and thus, unpredictable in any conceivable manner. There are two principal methods or types of random number generators (hereinafter called RNGs) used currently to generate random numbers, namely “true random number generators” (hereinafter referred to as TRNGs) and “pseudorandom number generators” (hereinafter referred to as PRNGs). TRNGs measure some physical phenomenon that is expected to have random characteristics and compensate for possible biases in the measurement process. Drawbacks to this approach include the need for measuring apparatus, and there are limitations to the rate at which random numbers (or “entropy”—a measure of unpredictability of information content) can be harvested since there is a dependency on the underlying rate of measurement. PRNGs entail the use of computational algorithms or functions to produce long sequences of apparently random results, which are actually completely predetermined by a smaller initial value, or seed. PRNGs tend to be fast, and act alone with the exception of input seeds which can be provided independently or are functions of their output. There are two distinct types of PRNGs—linear congruential generators which use a mathematical calculation and use outputs as a means of deriving inputs in a cyclic manner, and probability distribution generators which perform various operations on a data set or table (such as ARC4—a well-known stream cypher). Since both types of PRNGs use small seed or initialization values, there is a cryptographic quality issue in that that the entropy of the output number sequences with a PRNG is limited to the entropy of the input seed entropy. For example, if a numeral from 1 to 9 is used as a seed in a pseudorandom number generator, there may be a maximum of 10 variations of output, irrespective of the length of their output sequences. Therefore, PRNGs initialized using a seed value cannot satisfy the requirements of “information theoretic” security since their output keys are only truly random until they exceed the length of the input seed key.
It is therefore a specific objective of the current invention to address the random number generation problem through the manipulation of multiple RNGs or more specifically, the manipulation of the RNG internal states, operating within the cryptographic system.
Since constant reuse or use (in the case of stream ciphers) of the same key is cryptographically insecure due to the consumption of key entropy through normal use (resulting in a reduction in equivocation), in order to increase security, encryption key values may be frequently changed (with a penalty to bandwidth usage) to continuously add entropy to the cryptosystem, reduce the probability that an adversary may compromise the cryptosystem by intercepting and deciphering the encrypted messages, and thus requires new key values to be sent frequently by the sender to the receiver. In U.S. Pat. No. 5,412,730, Michael F. Jones “Encrypted Data Transmission System Employing Means for Randomly Altering the Encryption Key” we find such a method whereby new keys may be generated by a single independent RNG located at the sender, encrypted with the currently active key and transmitted along with normal data. Following decryption of the cipher-stream at the receiving station or entity, the encrypted keys are extracted from the data-stream, decrypted, and substituted for the prior key at a predetermined or designated point in time.
However, even such dynamic systems have numerous weaknesses, in that (a) pseudorandom keys are generated using PRNGs which eventually become predictable, (b) keys are stored and susceptible to interception and compromise, (c) keys are limited in length, and thus present a limit in cryptographic entropy and strength, even though resultant pseudo-key material is longer than the original key, (d) keys are reused, or transformed in a linear manner through functions, making equivocation analysis attacks viable, (e) the ciphertext is readily visible facilitating cryptanalysis attack and (f) the above issues lead to situations where the deciphering of a single key leads to compromise of successive and previous keys, and eventually to a compromise of all previous and future transmitted data.
In accordance with a principle feature of the present invention, multiple RNGs are deployed at both the sending and receiving stations or entities (may be users, machines, network devices, groups etc.) in that some RNGs are used to generate a sequence of random entropy values, which are mixed with message sequences, and communicated in an encrypted format, and are then used to alter the deterministic mechanics of the cryptosystem or it's RNGs in a probabilistic manner without actually communicating any keys, but merely the means of key generation. This mechanism therefore serves to “update” the entropy of the cryptosystem which ordinarily would be limited to the initial keys used, and will allow for an information theoretically secure cryptosystem, provided the rate of entropy updates exceeds the rate of entropy depletion. Whilst it is possible to pass new keys along in such a manner, or to pass updates which can be applied to existing keys to generate new keys, the rate of entropy depletion will usually be greater than the rate of entropy updates, which is why this technique is not generally used in cryptography. The solution to the problem of updating entropy at a faster rate than the depletion can be accomplished if a large entropy pool is used such as that found in a stateful random number generator. Merely passing the entropy updates between sender and receiver does not divulge any resultant keystreams or keys. In addition, provided an independent RNG is used to generate the entropy updates, and the entropy values are pre-encrypted with a shared key before being joined to normal messages or control values in randomly composed encryption blocks, and subsequently encrypted with the outputs of a second RNG, it is mathematically possible to augment the rate of entropy (or uncertainty) in the cryptosystem faster than it can be depleted through normal use. Doing this in a perpetually synchronized manner between sender and receiver, effectively results in a perpetual information theoretically secure cryptosystem, provided the proportion of entropy updates is larger than the amount of redundancy in the messages being communicated.
In accordance with the invention, in order to allow the two sending and receiving entities to continue communication using a random keystream, each is supplied in advance with a separate random number sequence which exclusively determines the numerical sequence and content of the random numbers produced by the two synchronized RNGs. Whilst the separate random number sequence may be used as a seed to restart the random number generation process at a predetermined time (as is found in prior art), it can alternatively be used to great advantage to alter a portion or the entirety of the output sequences of the RNGs, by altering a portion or the entirety of their internal states. Alternatively, it can be used for other purposes, such as allowing for variation in the encryption mechanics used, or for altering a portion or the entirety of the encrypted ciphertext prior to transmission.
In accordance with the principle feature of the invention, an additional independent entropy source like another RNG is used to generate random sequences of numbers of random length (entropy), which is then randomly and continuously injected into the data-stream or crypto-stream of stream ciphers, or into block ciphers, in order to randomly alter various components of the cryptosystem separately or in union, namely the output, the mechanics or the internal states of random number generators, or the mechanics or composition of the encryption algorithms, or the composition or characteristics of the plaintext prior to encryption, or the ciphertext after encryption prior to transmission. In order that the two generators switch from the previous cryptographic configuration to the next in synchronicity, means are employed at the sender to construct transmission packages which include data messages, random update information, other information and control information. Following encryption at the sender, transmission and decryption at the receiver, previous control information is used to deconstruct the package into its separate components and process data, updates, other information and control information accordingly. The construction and encryption, or deconstruction and decryption, of transmission packages (which may be fixed or randomly variable in length), may be referred to as an “encryption cycle”. Following the completion of an encryption cycle at the sender or the receiver, encryption configuration changes are applied to the respective cryptosystem components, and the new control information variables become active, replacing the old control variables. This continuous update process continues indefinitely until the communication link is terminated.
Without reiterating the above problems, it is therefore an object of the current invention to provide a method and a system which addresses the above problems whilst remaining within the confines of the characteristics of a “perfect secrecy” cryptosystem.
Until 1976, symmetric cryptosystems (that use a similar encryption and decryption key) like the one-time pad, were the predominant cryptosystem in practice. Unlike the one-time pad, most symmetric cryptosystems were block cyphers or stream ciphers, which used keys that were usually smaller than the message being transmitted. In the case of block ciphers the use of smaller repeating keys is not a cryptographically secure means of encryption, for two reasons, firstly it results in a decrease in the size of the cryptographic problem (by limiting the amount of cryptographic input entropy in a cryptosystem—the reason why “one time” pads should have keys as long as the message), and secondly, it allows for the implementation of equivocation techniques which use statistical means to compromise the cipher without having to resort to brute force techniques of trying every possible key (the reason why keys should never be repeated). Stream ciphers on the other hand use small keys as seeds to PRNGs for the generation of lengthy output keystreams, but the use of limited length seeds again results in a decrease of cryptographic entropy of the output key stream, the very reason why “perfect secrecy” systems require keys to be as long as the message. For stream ciphers or block ciphers to attain “perfect secrecy”, they must continuously receive random seeds, which results in 100% overhead to their message length and decreases their processing efficiency.
It is therefore a specific objective of the current invention to address the problem with ensuring that the one-time perpetual keys that are generated within the cryptosystem, can be initiated by a single set of initialisation values or keys, yet be able to maintain their unpredictable and perpetually random nature, through random feedback and synchronization updates between RNGs operating within the cryptographic system.
Like the one-time pad, all symmetric ciphers had the inherent “Distribution Problem” in that the original keys or seeds used for encryption and decryption must first be communicated between sender and receiver using a secure means of transfer. The security of any cryptographic system is only as secure as the initial key exchange, and the transfer process is prone to theft, interception, and alteration, defeating the purpose of the underlying cryptosystem.
The invention of asymmetric encryption (using different but related keys to encrypt and decrypt—such as private and public keys) addressed the distribution problem to a large extent allowing for limited length key distribution over insecure communication links, whilst also introducing authentication, verification and revocation properties which were previously absent in symmetric cryptosystems, an enormous achievement. However, this was achieved at the considerable expense of replacing previously mathematically proven “secure” (unbreakable) cryptosystems with a design alternative that relied on the unproven or assumed difficulty of solving certain mathematically “hard” problems.
It is therefore apparent that a need exists for a cryptosystem or cryptographic technique to address the issues that hampered the practical adoption of the one-time pad, by solving the randomness problem, the distribution problem, the key management problem, allowing for the use of asymmetric encryption advantages through the use of efficient super-encryption without compromising sound cryptographic principles by adhering to the basic requirements of a “perfect secrecy” cryptosystem. The present invention is directed toward providing such a technique and is unique in combining an autonomous network of multiple independent RNGs and an efficient super-encryption capability which is able to support asymmetric encryption solutions, as a viable means of addressing the problems mentioned above.
It is an object of the current invention to use the basic mechanics and principles of the OTP (true random numbers as long as the messages) as a foundation to the cryptosystem, and augmenting it where necessary with techniques to address past issues and keep all cryptosystem mechanics and functions dynamic, autonomous and independent from direct user interaction due to key input or message content. Thus ensuring that keys or RNG states are not readily available to user access since individual key values are automatically and dynamically generated, used and destroyed.
It is therefore an object of the current invention to provide a method and a system for the proper and cryptographically secure generation of unpredictable random numbers by dynamically and randomly synchronizing the state of independent random and pseudo-random number generators (RNGs) in an interconnected RNG communication network. The propagation of synchronicity entropy changes at a specific system node has the desirable side-effect in that its entropy impact may eventually be propagated across the entire network to other connected system nodes.
Another object of the current invention is to use the cryptographically secure output of multiple RNGs to perpetually and randomly add entropy to the encryption sequence in a manner similar to that found in prior art and referred to as “dynamic encryption” systems. Dynamic encryption attempts are described in U.S. Pat. No. 7,376,232, by Hamdy Soliman entitled “Computer System Security Via Dynamic Encryption”, and U.S. Pat. No. 5,412,730 by Michael F. Jones entitled “Encrypted Data Transmission System Employing Means for Randomly Altering the Encryption Keys”, but the present invention differs from previous art substantially in that it does not rely on the limited entropy of previous keys or data as a means of dynamic propagation, nor do keys get transferred over communication link, but in the transfer of an random length RNG state changes within a super-encrypted cipher-stream, which presents a more efficient and stochastic improvement, albeit at the price of expansion in message length, another reason why this technique does not appear in prior art, since message expansion is generally considered undesirable.
With regard to stream cyphers, ARC4, designed by Ron Rivest in 1987 and leaked to the public in 1994, is remarkable in its simplicity and speed in software. However, ARC4 has vulnerable weaknesses that argue against its use in new systems, such as not discarding the beginning of output streams, or when nonrandom keys are used. Additional problems identified by cryptanalysts with ARC4 include (a) a weak initial key scheduling algorithm, (b) a predictable starting distribution, (c) a uniform value distribution in that each value in the 256 byte secret state is unique and changed in location only, (d) a lack of additional input entropy following initialisation.
It is an object of this invention to introduce a viable and dynamic RNG (identified in the invention as a “Hi-Generator”) which addresses the problems identified with ARC4. Besides differences in the secret internal index pointers and state size, which can be randomly predetermined to be less or higher than 256 bytes, and can dynamically change its size during operation, it has no scheduling algorithm, nor a uniform distribution of unique values—all state values are random. In addition the pseudo-random generation algorithm does not merely relocate values, but amends them, in that each operation results in 3 or more additionally amended values. The most important difference being introduced with Hi-Generators, is that independent dynamic random changes are intermittently performed on the internal secret state of the RNG during normal operations, through the use of entropy updates. To conserve entropy, amended values need not be discarded but can be transferred to other Hi-Generators. These updates may be applied in an effort to synchronize variations in the output stream between synchronized RNGs, or to distribute the impact of entropy updates to unsynchronized RNGs.
A further object of the current invention is to simultaneously super-encrypt the combined data transmissions and entropy updates within a variable length message package, thereby using the RNG synchronization updates to “pollute” or “contaminate” the resulting cipher-text sequence for the message, making the cryptosystem resistant to linear and differential cryptanalysis, and removing the assumption that a message was even sent.
A further object of the current invention is to provide an efficient technique for the super-encryption of messages that incurs less than 100% overhead in message length, and which is able to shield the cipher-text from cryptographic analysis and address the “magic number” and “null string” problems.
A further object of the current encryption is to allow for the integration of existing authentication and verification asymmetric encryption mechanisms (used for user authentication, identification information, keys or certificates) into the encryption process itself, greatly adding to the security of asymmetric encryption systems and extending the present invention cryptosystem to include central authority control and management functions, and supporting the distribution and authentication of cryptosystem entities (users, machines, systems, data, access control and services etc.), the establishment of secure communication links between vetted entities, and eliminate the possibility of unidentified entities communicating to any other entity within in the cryptosystem network.
A further object of the invention is to provide a single efficient, fast and scalable means of encryption which can be used for all eventualities, namely encryption of data at rest, data in transit over communication channels, and for the provision of a means of authenticating network entities.
A further object of the invention is to provide a means for ensuring that all traffic flowing within the cryptosystem network remains behind a logical security boundary, whereby access to and from specific system nodes is controlled identified and kept under user or management control.
These and other objects, features and advantages of the invention will be apparent to those skilled in the art, and more clearly understood by considering the following detailed description of specific embodiments of the invention.
BRIEF SUMMARY OF THE INVENTIONThe purpose of this invention is to provide a secure cryptographic method and system capable of securing the transfer of data within and between computer systems, whilst addressing the problems of pseudo-random number generation, infinite length keystreams and efficient super-encryption.
To this end, the cryptosystem uses multiple RNGs (preferably of a type known as Hi-Generators) to produce multiple random keystreams which are used, for the encryption (preferably using the Hi-Encrypt super-encryption algorithm) of normal data transmissions and the transfer of random entropy updates to other system entities (nodes, users, data, RNGs, machines, devices, computers, etc.).
In so doing the invention addresses the following problems: (a) the difficulty in producing unpredictable infinite length random number sequences which are at least as long as the messages which need to be transmitted, (b) protecting the cipher-text through multiple encryption, from most cryptographic attack vectors, including brute-force attacks through the use of a non-deterministic encryption method which allows for “perfect secrecy” and results in multiple false positives, (c) providing a secure means of secret key generation, distribution, use, disposal and storage to prevent key interception, (d) providing a secure means of initial key distribution without exposing public or private key contents to unauthorized parties, (e) allowing for central authority management of network entities, (f) preventing unauthorized parties from interfering with or intercepting dynamic keys through replay or man-in-the-middle attacks, since no actual keys are transmitted over communication links and interference with the key generation mechanism ultimately leads to intrusion detection, and resynchronization of the specifically affected encrypted channel.
The use of Hi-Generate RNGs and the Hi-Encrypt super-encryption algorithm results in a cryptosystem with both “perfect secrecy” and “ideal secrecy” characteristics.
This and other features and advantages of the invention may be more clearly understood by considering the following detailed description of specific embodiments of the invention.
In the course of the brief and detailed description to follow, reference will be made to the following attached drawings, in which:
Some embodiments described herein are associated with “encryption” and “decryption”. As used herein, the phrase “encryption” is the process of encoding or transforming messages or information in such a way that only authorized parties can transform it into a meaningful message. Encryption does not prevent interception, but denies the message content to the interceptor (as published by Goldreich, Oded “Foundations of Cryptography: Vol 2, Basic Applications. Vol. 2”, Cambridge University Press, 2004). In an encryption scheme, the message or information, referred to as “plaintext” is encrypted using an encryption algorithm, producing “ciphertext” that can only be read if decrypted. For technical reasons, the security of encryption schemes are usually dependent on true random or pseudo-random encryption keys generated by an algorithm. It is in principle possible to decrypt the message without possessing the key, but, for a well-designed encryption scheme, even infinite resources, skill and time are insufficient to decrypt the message to a single correct message. Such a cryptosystem is said to have the characteristic of “ideal secrecy”. In addition, “perfect secrecy” is a characteristic of a cryptosystem in that, given an encrypted message (or ciphertext) from a perfectly secure encryption system (or cipher), absolutely nothing about the unencrypted ciphertext (the plaintext) will be revealed in the ciphertext. Technically speaking, a cryptosystem has perfect secrecy if for any message “x” and any encipherment “y”, p(x/y)=p(x). This implies that there must be for any message/cipher pair, at least one key which connects them.
Some embodiments described herein are associated with a “network entity”, “system entity” or “entity”. As used herein, the phrases “network entity”, “system entity” or “entity” may be used interchangeably and may refer to any physical or logical component or element of the cryptosystem that uses, interacts with or constitutes said cryptosystem and which can be separately identified and authenticated using an identification number sequence. Examples of entities include but are not limited to users, machines, groups, enterprises, authorities, services, ports, keys, RNGs, entropy sequences, programs, addresses, tokens, devices, systems, configurations, components, nodes, instances, states, identifier types, ciphers, keystreams and data streams. Entities may comprise one or more other entities.
Some embodiments described herein are associated with a “network node” or “communication node”. As used herein, the phrases “network node” or “communication node” may be used interchangeably and may refer to any physical or logical device that can communicate via a network, to another network node, or to itself, or to another network component within the same network node. Examples of network nodes include a Personal Computer (PC), a workstation, a server, a printer, a scanner, a facsimile machine, a copier, a Personal Digital Assistant (PDA), a storage device (e.g., a disk drive), a hub, a router, a switch, and a communication device (e.g., a modem, a wireless phone, etc.). Network nodes may comprise one or more network components. As used herein, the term “network component” may refer to a network device, or a component, piece, portion, or combination of network nodes. Examples of network components may include a Static Random Access Memory (SRAM) device or module, a processor, and a communication path, connection, port, or cable.
In addition, some embodiments are associated with a “network” or a “communication network”. As used herein, the terms “network” and “communication network” may be used interchangeably and may refer to any object, entity, component, device, and/or any combination thereof that permits, facilitates, and/or otherwise contributes to or is associated with the transmission of packets, sequences, signals, and/or other forms of information between and/or within one or more network devices. Networks may be constituted of, or include a plurality of interconnected network devices. In some embodiments, networks may be virtual, neural, artificial, hard-wired, wireless, and/or of any other configuration and/or type that is or may become known. Communication networks may include, for example, one or more networks configured to operate in accordance with the Ethernet LAN transmission standards published by the Institute of Electrical and Electronics Engineers (IEEE).
Further, some embodiments herein are associated with “data” or “information”. As used herein, the terms “data” or “information” may be used interchangeably and may refer to any individual or combination of sequence, numbers, data, text, voice, video, image, message, bit, byte, packet, package, pulse, tone, waveform, and/or other type or configuration of signal and/or information. Information may be or include information packets, network management information, cryptosystem control information, datagrams, or cryptograms transmitted, for example, in accordance with the Internet Protocol Version 6 (IPv6) standard as defined by “Internet Protocol Version 6 (IPv6) Specification” RFC 1883, published by the Internet Engineering Task Force (IETF), Network Working Group, S. Deering et al. (December 1995). Information may, according to some embodiments, be compressed, encrypted, checked for errors, certified, and/or otherwise packaged or manipulated in accordance with any method that is or becomes known.
Some embodiments described herein are associated with “information indicative of” or “indicia” of information associated with a subject, item, entity, and/or other object and/or idea. As used herein, the phrases “information indicative of” and “indicia” may be used to refer to any information that represents, describes, and/or is otherwise associated with a related entity, subject, or object. Indicia of information may include, for example, a code, a reference, a link, a signal, an identifier, and/or any combination thereof and/or any other informative representation associated with the information. In some embodiments, indicia of information (or indicative of the information) may be or include the information itself and/or any portion or component of the information.
In addition, some embodiments are associated with the protection, encryption and utilization of information against “risk” or “security risk”. As used herein, the term “risk” or “security risk” may generally refer to any probability, possibility, chance, and/or likelihood of incurring an event of negative consequence such as encountering peril, loss, hazard, danger, and/or injury. In some embodiments, “security risk” or “risk” may refer to any actual, perceived, inherent, intrinsic, and/or other form or type of peril, loss, hazard, danger, and/or injury. According to some embodiments, “security risk” or “risk” may refer to any individual, item, device, event, agent, organization, and/or other object or entity that may contribute to, cause, result from, and/or otherwise be associated with any type or form of risk. Risk may, according to some embodiments, be associated with any type and/or aspect of possibility of peril, loss, hazard, danger, and/or injury. Examples of risk may include, but are not limited to, financial risk, political risk, risk of injury, risk of theft, risk of damage, risk of terrorism, risk to reputation, regulatory risk, legal risk, war risk, and environmental risk.
Upon reading this disclosure, those skilled in the art will recognize that embodiments of the present invention may be used to encrypt, decrypt, authorise, authenticate, verify, repudiate, propagate, alter and any number of different types of information.
SUMMARY OF THE INVENTIONThe following descriptions of embodiments, some of which are preferred, are provided for further illustrating, but not limiting, the present invention.
To alleviate the problems inherent in prior art, embodiments provide systems, methods, apparatus and means for random number generation, encryption, decryption, super-encryption, super-decryption, authentication, message packet construction and message verification. The applicants have recognized that there is a need for separate systems, methods, apparatus and means for addressing specific cryptographic problems, and a need for the integration of those separate systems, methods apparatus and means to be used in unison to create a fully integrated cryptosystem which strives to attain the security characteristics of “perfect secrecy”, or “ideal secrecy.
One benefit of some embodiments is that the cryptographic security of encryption through the use of multiple random and perpetual keystream sources, addresses a number of issues with current computer security implementations in one comprehensive solution. As a result, users such as financial institutions, businesses, government agencies, or the like may be afforded a greater level of security in protecting their digital assets, such as files, data, network communications, in perpetuity. Some embodiments provide benefits such as the ability to perform a number of cryptographic functions, such as unilateral initiation of communication, encryption to local media, encryption over a network, authentication, verification, digital signing, hashing, broadcasting, automated key management, stream auto correction, and others all using a single cryptographic algorithm or cipher. Other features and advantages that are derived from some embodiments will become apparent upon reading this disclosure.
Individual Cryptographic System ComponentsCertain embodiments of the current invention include, but are not limited to the use of a number of separate components, modules, methods, systems, apparatus and means which may in themselves be considered separate inventions. The separate methods, systems, apparatus or means which will be discussed include: (a) an random number generation method or system used within random number generators having variable and random internal states, to produce random number sequences (b) a method or system of deploying multiple independent random number generation sources in various configurations using a hierarchy, making them conducive to another mechanism detailed below which allows for the generation of unpredictable random number sequences using pseudo-random number generators (c) a fast, simple and efficient super-encryption method or system using two or more ciphers, and one or more keystreams, which is brute force resistant, (d) a message packaging method or system for the transmission of data and additional random entropy sequences using communication links, (e) a method or system of using the additional entropy information to alter the dynamics of the cryptosystem at various levels, (f) a method or system for the distribution of initial keys or sequences to initialize secure communication links between network nodes using simple delivery, by integrating asymmetric encryption components within the super-encrypted cryptogram (g) a method or system of centrally registering cryptosystem entities (users, machines, keys, algorithms, devices etc.), (h) a method or system of exchanging, distributing and protecting initialization keys or sequences between cryptosystem entities.
Random Number GenerationCertain embodiments of the current invention include, but are not limited to the use of a distinct and unique random number generation method or mechanism called “Hi-Generate” which can be used with PRNGs which use secret internal states to generate a pseudorandom or random stream of values in bits or bytes. Whilst PRNG's used may use internal secret states of any length, the Hi-Generate method requires these states to be composed of random numbers and not a uniform random distribution of unique numbers as is encountered with ARC4, a popular stream cipher.
Like any stream cipher the sequence of random numbers produced by a Hi-Generator can be used as a keystream for encryption purposes by combining it with plaintext using a bit-wise exclusive-or (XOR), or logical operation, with decryption being performed using a similar operation with XORs (since XOR with given data entails an involution), or by reversing the logical operation. To this end it is similar to the “one-time” pad with the exception that a generated stream of pseudorandom bits, rather than a readily available pad, is used. The output random stream sequence of numbers can also be used to great advantage for a number of alternative purposes under specific conditions which will be become apparent as we discuss the invention in more detail.
With reference to
A Hi-Generator may take many different configurations which may include additional 8-bit or other bit-sized index pointers 108 per state array, or additional internal state arrays 109 with associated index pointers for each state array, or even additional independent Hi-Generator secret states 110, containing their own separate state arrays and index pointers. In addition, additional random number outputs 111, 112, can be derived from multiple arrays within a secret state, from multiple secret states within a generator, or multiple output bytes can be derived from a single state 113. Since state array lengths may easily exceed 256 characters, index pointers may be 16-bits in size or larger where this is the case. There are usually at least two index pointers per array, but in some embodiments, three or more may be deployed.
In describing the general internal mechanics of a Hi-Generate RNG, such as initialization, or internal operations of a permutation, we may use a basic example such as that shown in
The initialization process of a Hi-Generator or other RNG may take many forms and generally entails a fixed starting configuration of the RNG prior to commencing random number generation. Certain embodiments of the current invention include but are not limited to deployments where the initial composition of RNGs is established using a predetermined, provided, fixed or random sequence which may be numbers, data or any other type of information, such as a key Such a process may designate the number of state arrays, appropriately sized index pointers, their relationships, and their respective content with regard to values. Generators may therefore be initialized with any randomly determined or predetermined length, and appropriately sized index pointers, and can simply be initialized by filling any arrays or index pointers with the provided random or predetermined number sequences. Where provided sequences are shorter in length than array A such as with a key, simply repeating the sequence ad nauseum should suffice. Alternatively, the content of another generator may be used, with or without a transformation.
It is important to make a distinction at this stage between synchronized and unsynchronized RNGs. Whilst both unsynchronized and synchronized RNGs may share all or a portion of internal information with other RNGs, unsynchronized RNGs do not produce or intend to produce identical sequences with other RNGs. Synchronized RNGS use the content of another RNG, without a transformation, such that the sequence of random numbers may be identical, allowing for encryption by a sender and decryption by one or more receivers. Hi-Generators or any other PRNGs for that matter, require that the sender and receiver Hi-Generator or RNG configurations be perfectly synchronized in composition and array values. Unsynchronized RNGs on the other hand operate independently. Whilst this invention provides for additional secure means and methods of initializing synchronizing RNGs, which will be discussed later, we will simply state at this point that the sending generator has communicated the details of an identical Hi-Generator or RNG configuration to the receiving station and with reference to
Certain embodiments of the current invention include, but are not limited to the random number generation of single values (which may be of any bit size) per generator iteration and may generate multiple values per iteration to one or more output sequences. With reference to
The pseudo code for single value generation over one iteration is:
Certain embodiments of the current invention include, but are not limited to the random number generation of multiple output values per generator iteration. With regards to the specific mechanics of the Hi-Generator method for multiple value outputs we see from
The pseudo code for multiple output number generation over one iteration is:
Alternatively, in order to increase parallelism, single or multiple output sequences can be provided by using multiple single-value or multi-value modes of operation with a Hi-Generator, or multiple Hi-Generators. A Hi-Generator is in essence an autonomous software or hardware unit that is used to produce random numbers. Hi-Generators may be internally composed of one or more secret internal states, each having its own array, index pointer, and output configuration. A number of Hi-Generators running at a specific logical or physical location may also be grouped together as an autonomous unit and can be referred to as a RNG node or Hi-Generator node. Hi-Generate mechanics can be applied to other RNGs, provided they have large internal states with random values.
In addition, it may be possible to synchronously or asynchronously apply multiple Hi-Generate random number generation procedures simultaneously to the same state array, and still produce separate random number sequences. Alternatively, in order to avoid collisions, a large state array may be segregated into sections which may be of variable length with one or more random number procedures being uniquely or in combination be applied to one or more sections.
Certain embodiments of the current invention include, but are not limited to where a sender and a receiver on opposite ends of a communication link may use multiple Hi-Generators to initiate communications. The sender and the receiver may use an initialization process which may entail the use, interaction or combination of and between two or more Hi-Generators or RNG sources, each of which may consist of one or more RNG arrays, secret states, instances or nodes. From
The key may be applied literally to any other initialisation package or additional key, or it may be used to introduce random effects on RNG mechanics, such as altering specific values in RNG states, or entail counts where data is continually applied to the internal state, arrays or index pointers of an RNG during initialisation until a predetermined insertion or output count is attained. In addition, the input or output sequences to and from any initialized or uninitialised generator may be intermittently interrupted during initial communication sessions, or during initialization activities, using a predetermined characteristic of the key or a function of the key. For example, it may miss 20 values, output 34, then miss 200, then output 7, etc. until RNGs are fully populated or a predetermined count is reached. After that point, the respective RNG may be synchronized with its counterpart and output may be used to initiate secure communication using RNG keystreams, or additional initialisation activities, or both
In addition, the process can be reversed, using a predetermined instance configuration, after RNG B has been initialized, to reinitialize RNG a generators in a similar manner, in order to have one or more additional arrays initialized and synchronized with another Hi-Generate instance. Thus it is possible to have a number of sending and receiving keystreams that have been intialised and synchronized with receiving and sending keystreams of another Hi-Generator at the other end of a secure communication link.
In addition, the above initialisation process may be used to initialize multiple Hi-Gen instances, or further, to initialize Hi-Gen nodes and maintain synchronicity with other Hi-Generate instance and nodes, prior to initiating keystreams.
In addition, the process mentioned above can be used to prepopulate backup Hi-Generators or their respective internal states, in case of communication errors in a secure communication link. For instance, once RNG B generators at both locations have been initialized, RNG A generators may have their internal secret states saved prior to reinitialisation, with or without an advancement count. An advancement count allows an RNG to produce output sequences intermittently, in that specific values or value ranges from the output of an RNG may be included or excluded from the output stream of random numbers, dependent on a predetermined characteristic of the key, or the initialisation package.
In addition, once the first initialization package has been processed and a secure communication link has been established between the secondary RNG A generators at both locations, any RNG at both locations may be used to generate a new initialisation package which can then be appropriately deployed at both locations on any existing RNG or any newly deployed RNG, and this initialisation process can be repeated ad nauseum until any predetermined characteristic has been satisfied.
Random Number Generator Updates and FeedbackCertain embodiments of the current invention include, but are not limited to situations where multiple RNGs having various configurations and internal secret states may randomly or intermittently exchange entropy between each other, which may be used to great advantage to introduce random and unknown characteristics and impacts into the random number production mechanics resulting in unpredictable random number sequences being generated using pseudo-random number generators.
Considering the situation of three RNGs, it may be possible for a first RNG to produce a random sequence, which may then be encrypted or processed with the random sequence of the second RNG. The resultant random number sequence (in effect a ciphertext) may be used as a random sequence in its own right, or it may be transmitted to another third RNG which has been synchronized with the second RNG and decrypted, making the initial random sequence produced by the first RNG available at the location of the third RNG. The available random sequence may then be used to effect changes amongst other RNGs at the third RNGs location.
In the case of two simple RNGs with any number of internal secret states, arrays and associated pointers, a first (or primary) RNG may be used, under predetermined or random conditions, or following a predetermined characteristic of a generated output sequence of numbers, to interfere with the internal secret state (arrays, index pointers, configuration and values) of another second (or secondary) RNG by updating or altering predetermined or random sections of specific or randomly selected internal secret state components, with the mentioned generated output sequence. The scope of the updates or alterations includes the selection of RNGs, and alterations or updates may be determined by predetermined characteristics of the output sequence, such as determining the order of arrays to be updated using a byte from the output sequence and applying a modulus of the number of arrays. In addition, the specific updates or alterations which may be effected, may be predetermined, or may be derived from byte values in the output sequence, for example the 2nd 3rd and 4th byte may replace some or all index counters, the 5th byte may indicate a start position for the update in the array for the length of the output, or the 6th byte may indicate the end index position of the subsequent bytes to be applied. Changes, updates and alterations may be applied to the designated RNGs internal state arrays and index pointers directly, using a function or transformation, or through encryption with a secret key. In addition, alterations and changes may have feedback or feed-forward effects in that affected or updated values in the second RNG's state, arrays or index pointers may be extracted prior to being amended and be sent back to the first RNG, or be forwarded to another RNG which may be local or remote. This forward update and feedback process may also be subject to weighted conditions of the input, resulting in a rudimentary implementation of an artificial neural random number generation network. This update, feedback and feedforward procedure may be initiated by any RNG, through any RNG, to any RNG.
With a view to an artificial neural RNG network, in order to avoid the problem of convergence (which is wholly unwanted in a random number generator and entails a biased trend in the reduction of entropy in the network), and step beyond pseudorandom number generation into the realm of true random number generation, it is necessary that the composition, timing, or means of application of the update, is subject to some random unpredictable interference (entropy) originating from an external true random source, such as user keyboard timings, network timings, noise channels etc. Since update and feedback routines may also occur between any combination of state arrays, instances or nodes, another means by which one can guard against the artificial neural network “convergence” problem, is to ensure that random number generators are deployed using a one-to-many hierarchy of primary to secondary generators. Primary RNGs serve as rendezvous points for feedback or feedforward updates from secondary RNGs. From a scalability perspective, multiple primary RNGs may be deployed with a master RNG above them. Primary RNGs are sources for update propagation down to secondary RNGs which may transfer updates to other network node secondary RNGs, which may then be propagated as feedforward up to their primary RNGs. In this manner, small updates which implement random changes to RNG internal states may result in mass propagation throughout the network. Having additional true random sources of entropy on even a single network node should be sufficient to prevent convergence throughout the network, considering that amount of propagation which may take place in a cryptosystem.
In addition, since secondary RNGs may receive updates from local primary RNGs and transfer these updates across the network to remote secondary RNGs, the updates may be applied to the local secondary RNG immediately after transmission and to the remote secondary RNGs immediately after receipt and decryption, in order to maintain a synchronized secure communication link. Updates to local secondary RNGs may result in feedback changes being pushed back to local primary RNGs, identical updates to remote synchronized secondary RNGs may result in identical feedforward updates being applied to remote primary RNGs. Whilst remote and local primary RNGs may receive similar feedback/feedforward updates, their existing internal states may be different resulting in updates having different impacts. For this reason, updates should be applied using an index retrieved from the primary RNG. In order to avoid convergence, the update point of the target RNG array—let's call it A[i]—may be determined by a predetermined characteristic of the update sequence. A logical operation on the value of “A[i]” and “I” (modular the size of the target array) may be used to determine the actual insertion point of the feedback/feedforward update.
As can be seen, update impacts eventually propagate to the entire network since they are escalated up the RNG hierarchy of other systems, introducing new entropy into the primary RNGs, which then propagate downward again and are transferred to other systems. This mechanism ensures that the impact to random number generation sequences from a single update sequence eventually permeates throughout the entire network, and continues to do so long after the initial update sending station has been closed down.
Entropy InjectionCertain embodiments of the current invention include, but are not limited to embodiments where random number updates and feedback to and from generators as mentioned above may serve as a means of introducing additional entropy into the random number generation process of other RNGs and therefore into the encryption process itself. From
In a similar manner, it is possible to combine random number sequence updates from one or more random number generator sources and transfer them to one or more random number generator destinations. The means, manner and method in which random number sequences may be combined may be predetermined by cryptosystem configuration, or by predetermined characteristics of the update information being transferred, or predetermined by characteristics of the originating RNGs and destination RNGs of the random entropy update information. For example, it may be that separate random number updates from a primary RNG (update 1) and secondary PRNG (update 2) at station 1 may be appended (update 1-update 2) or multiplexed prior to transmission to a primary RNG and secondary RNG at station2, where upon receipt they (update 1-update 2) are split or de-multiplexed. Following transmission, “update 1” may be applied to the secondary RNG at station 1, upon receipt at station 2 it may be also applied to the secondary RNG at station 2. Upon receipt at station 1, “update 2” may be applied to the primary RNG at station 2. Alternatively, entropy update(s) originating at one or more RNG(s) may be transmitted in any form and be applied to one or more other RNG(s), in whole or in part, or in combination with any other data or information.
Entropy Encryption and TransferCertain embodiments of the current invention include, but are not limited to embodiments where random entropy updates are encrypted or super-encrypted before being transmitted to a destination. This may include scenarios where one or more random entropy updates or keystreams originating from two or more RNGs are encrypted or super-encrypted with one or more keystreams from the same RNGs, or one or more keystreams from two or more other RNGs. With
In this manner, entropy updates can be cascaded down the RNG hierarchy of a local or remote system, transmitted in an encrypted form, and be cascaded up the RNG hierarchy of another local or remote system. Thus, through a single efficient action, two synchronized RNGs can be synchronously subjected to a “random walk”, and enable the propagation of an entropy update on another system as feedback. This encrypted entropy transfer also serves another advantageous purpose, in that the entropy updates can introduce considerable “cryptographic pollution” when encrypted entropy updates are combined with other data transmissions.
Encryption and Super-EncryptionIn order to achieve efficient super-encryption, two or more different encryption sequences must be used. Whilst it is possible to super-encrypt a message with a equal length keystream, encrypt the keystream with a block cipher using a short key, and combining both ciphertexts, this is inefficient, since it doubles the overall message length—100% overhead. Certain embodiments of the current invention include, but are not limited to embodiments where super-encryption can be attained with considerably less overhead than 100%. Since random entropy updates may be included with the data to be transmitted, and may be as large as the message, this is also not efficient. Whilst having the advantage of introducing additional entropy with every encryption cycle, increasing cryptographic strength, entropy updates of a fraction of the message size may be used, if they are applied at random to their destinations, using random numbers. In addition, whilst ciphertext may be transmitted in uniform length blocks from one station to another, the underlying message packages may be of random length, with randomly sized data and entropy update populations, and may be subject to random variations in encryption cycles. The current invention may use efficient deterministic encryption algorithms such as that shown in
In accordance with another principal feature of the invention, the assembled message transmission package which may contain message, random numbers and cryptosystem management information may be encrypted using two or more independent and different encryption algorithms using the output of one or more RNGs, and decrypted in a reverse manner. Ideally, a value at a specific location in a plaintext should undergo a value and a location transposition, and thus the need for two or more ciphers during super-encryption. The overhead added to the packet is equal to the amount of random update material of random length that has been added to the message package, and should be usually between 0% and 25% for an average of 12,5%. In keeping with the principles of “perfect secrecy”, the use of such a super-encryption mechanism, along with the use of a randomly determined message transmission package size which may have its last package size appropriately amended to appear as if it is using standard length processing blocks, may ensure that no information regarding the message, including whether a message was even transmitted, may be leaked to an assailant. Even in the event of an assailant being in possession of one or more known plaintext/ciphertext pairs, it may be considerably difficult, if not impossible for the assailant to reconstruct the underlying RNG states using a brute force approach, since updates may occupy entire message packages alone, and their impacts which may or may not affect future RNG outputs may eventually be erased completely.
Certain embodiments of the current invention include, but are not limited to embodiments where the encryption or super-encryption process used, which may be called “Hi-Encrypt”, may use an XOR and a shuffle, or a combination of multiple XOR and shuffle encryption steps, using key sequences or portions of key sequences provided by separate or similar RNG keystream sources. In addition, when message or transmission packages (containing messages, RNG updates, control information and keys) exceed 700 bytes, there are more combinations of shuffling a sequence, than the information that can be contained in that sequence. In addition, when using large message package sizes in excess of 256 bytes, it may be feasible to use a single shuffle with 16-bit counters, or multiple shuffle encryption steps. XORs may be conducted before shuffles or vice versa. A combination of modular addition, subtraction, XORs and Hi-Shuffles may also be used in any viable combination, such as “Hi-Shuffle/XOR/Hi-Shuffle”, or “Hi-Shuffle/XOR/XOR”, or “Hi-Shuffle/Hi-Shuffle/XOR”. Naturally, when decrypting, the algorithms are conducted in reverse.
Certain embodiments of the current invention include, but are not limited to embodiments where 2 or more parallel Hi-Encrypt encryption steps may be used to encrypt a sequence with an “XOR/Hi-Shuffle” mechanism, conducting the XOR first, followed by the Hi-Shuffle, or vice versa, and XORing and Hi-Shuffling the plaintext in one single step, using the 2 keystreams. Alternatively, 2 simultaneous XORs and a shuffle may be performed. The decryption procedure would be a logical reverse of operations.
Certain embodiments of the current invention include, but are not limited to embodiments where a Hi-Generate method of keystream generation and a Hi-Encrypt method of super-encryption results in no keys being stored on any system, keys being immediately used and destroyed, whilst eroding the underlying structures used to generate such a key. For the purposes of key destruction or destruction of sensitive data, the Hi-Generate and Hi-Encrypt process can be effectively used to rapidly erase or overwrite any sensitive information, using very large state arrays or RNG internal configurations (multiple variable interacting state arrays or instances).
An issue with current asymmetric encryption is that public keys are visible during transmission and can be immediately attacked. The encryption process can include the application of a Diffie-Hellman asymmetric encryption scheme, which is used to apply public/private key operations to the RNG updates themselves, within the shielded confines of the super-encrypted package. In order to exorcise man in the middle attacks, it is possible to use an interlink protocol, where parties send partial message updates to each other. Since initial communication packages are essentially acknowledgments, there is considerable empty space for large initial random updates. Following a predetermined synchronisation from a known initial PRNG state provided by the sender, any intermediary may be able to decrypt messages sent by sender, however, replies by the receiver may be more difficult to keep track of, since public asymmetric encryption may obscure the PRNG updates continuously. With sufficient initial synchronisation and random intermittent asymmetric operations, an adversary may quickly find himself overwhelmed by the computational mass increased exponentially by additional encrypted PRNG updates. The authentication mechanism essentially provides a better means of solving the distribution problem without revealing public keys. Additional elements may be included to incorporate a dynamic element to the Diffie-Hellman sequence, or to use it for specific multi-factor requirements which may present themselves over time in the future.
Certain embodiments of the current invention include, but are not limited to embodiments where random number updates or any other information, including encrypted crypto-streams or ciphertext may be advantageously subjected to a pre-encryption process, for example with asymmetrical encryption public keys and private keys, prior to being inserted into the main super-encryption process, as a means of inserting authentication, verification or key information within the protective confines of the super-encryption process, and allowing secure transfer until extraction from the crypto-stream and subsequent decryption via an asymmetric key at the receiving end. The use of such an asymmetric key pre-encryption sequence may have the advantage that a sender can transfer an initial RNG setup and one of a set of asymmetric keys to a receiver who sets up an identical RNG configuration. Whilst communication from sender to receiver may be trivial to decrypt, the reverse process from receiver to sender poses an ever increasing cryptographic problem. By way of an example with three parties, “Party A” (a sender at Station 1), “Party B” (a receiver at Station 2) and “Party I” (an intruder), we see that Party A may send Party B an initialisation sequence (composed of a sequence of random numbers, and a public key) which is intercepted by Party I. Whilst Party I may be able to easily decrypt any further messages sent by Party A to Party B based on the initialisation sequence which may be used to synchronise the receiver's RNG with the senders's RNG. However, Party B may initialize his RNG with the RNG update and initiate a series of RNG updates which are super-encrypted with the current state of his RNG and include update material encrypted with the public key, from his location to Party A's RNG. This may require Party I to break the public/private key sequence for every encryption cycle in order to access the update material for the next super-encrypted encryption cycle. This process can be repeated ad nauseum for a predetermined number of cycles. In addition, RNG updates may also include information which applies a portion of the RNG update and advances the RNG generator, before applying other portions.
In addition to inserting asymmetrically encrypted RNG update information into the Hi-Encrypt message package prior to super-encryption in order to initialize an RNG pair, Party A and Party B may also rely on a method to address man-in-the-middle attacks, by each party simultaneously sending the other an independent RNG update sequence, but splitting the asymmetrically encrypted RNG update information into two or more Hi-Encrypt message packages, and where each party alternates by sending the other one of the required packages in an alternating manner, similar to that used with an “interlink protocol”. This allows for each party to separately initialize an RNG pair between them, and thereby initialize 2 separate communication channels. The method demonstrated above may also be used during normal operations to prevent “man-in-the-middle” attacks.
This RNG initialisation and synchronization procedure may be used indefinitely as part of the standard encryption operation, or it may be randomly introduced at predetermined points in time, or it may be used to exchange recovery information which may be used to automatically reinitiate communication links under predetermined conditions.
Certain embodiments of the current invention include, but are not limited to embodiments where data or information must be encrypted locally. In the case of where local encryption is required, such as encrypting local keys, encrypting local data, backup RNG updates or RNG configurations, it is important to ensure that decryption keys are not readily available locally. In such a case, a separate key management process may be used at a station for the purpose of key management, in that information indicative of the local entity which requires encryption (userID, password, machineID, processID, serviceID etc.) may be converted to an initial keystream or key, which may be used to Hi-Encrypt a random number sequence (or vice versa) into a second keystream or key. The ensuing keystream or key is then used to Hi-Encrypt the information or data which needs to be secured. The secured and encrypted data may then be stored with an identifier, and the initial random sequence. Any information which relates specifically or is indicative of a specific entity (user, machine, process, address, permissions, RNGs, links etc.) may be Hi-Encrypted in such a manner.
Certain embodiments of the current invention include, but are not limited to embodiments where Hi-Encryption of messages, information, random sequences, random updates, random feedback or any other information may be conducted with any other existing encryption algorithms or combinations of encryption algorithms, including but not limited to asymmetric and symmetric encryption algorithms, ciphers or cryptosystems. Multiple encryption and decryption algorithms may be applied in any order during encryption and decryption. For example, random updates or feedback may be encrypted or super-encrypted with an asymmetric public or private key pair shared between the sender and the receiver, prior to being inserted into a data transmission stream package of variable length (containing message data, encrypted random updates, control information and additional symmetric and asymmetric keys), which is then subjected to a super-encryption process prior to transmission of the crypto-stream. At the receiving end, the crypto-stream can be decrypted or super-decrypted, whereupon the data, control information and the encrypted random update can be extracted. The encrypted random updates can then be decrypted at the receiving end and applied according to an agreed predetermined characteristic. In addition, the application of random updates may include the update being applied to any number of synchronized or unsynchronized RNGs at any of the multitudes or plurality of sending and receiving stations. In addition, RNG update sequences may also include the transfer of new asymmetric or symmetric keys in that RNG update sequences may contain such key values. In addition, RNG update sequences may be transferred to other parties and then be subject to amendment by symmetric or asymmetric keys, which may accompany the updates by way of encrypting, converting, appending, inserting, or replacing the key material with random update information, or vice versa, where random update information is encrypted, converted, inserted, appended or replaced by key information.
Message PackagingCertain embodiments of the current invention include, but are not limited to embodiments where one or more single or combined sequences of information indicative of data, or indicative of RNG update information, or indicative of specific characteristics of an entity (user, machine, program, service, port, address, channel, device, or system, authentication, identification), or indicative of management and control information required for effective management of cryptosystem activities or actions, is encrypted or super-encrypted using one or more keystreams. This may include combining or packaging message data, control information and random update information from one or more source PRNGs, in any combination, into a single or multiple encryption processing unit or message package, and encrypting or super-encrypting the package using one or more keystreams from one or more RNG source. The resulting cypher-stream may be used for further cryptographic purposes such as initialisation vectors, RNG updates, user and entity authentication, or may be stored locally for encryption of data at rest, or may be transmitted to remote locations.
In addition, the invention may use variable length message packages of any size (usually between 100 and 1536 bytes), taking into consideration resource consumption, and which may be composed of various variable length compartments for control information, messages, RNG updates or random sequences, and cryptographic management information. The primary compartment, namely the package control information compartment, may contain all information relevant to the package or the message, namely compartment sizes, message authentication digest information, total message size to be transmitted, and a random number which is used to obfuscate such information. The message compartment may contain the message or a portion of the message being transmitted. The random information compartment contains random entropy sequences such as RNG updates. The management compartment contains information which may establish predetermined conditions to be applied in the next package sequence. Package control information may usually be provided in bytes and may be about 10 to 20 bytes in length (4 bytes for total message size, 4 bytes for Message digest, 1 random byte, 2×3 bytes for compartment sizes, 2 bytes for a random sequence number and 1 byte for a package type. Package types may designate 256 different packages, RNG configurations, or encryption/decryption configurations and methods. Management information may be about 2 to 10 bytes long and is used to control the application of random entropy updates. It may indicate RNG sequences, index positions for application of information, RNG cycles to skip etc. Alternatively, it may use 2 bytes (or more) to identify any of 16535 preconfigured activities to be used in the next encryption cycle. Message information and random update compartments may be managed as a single unit or determined randomly. In order to ensure that super-encryption can be conducted with a payload less than 100%, it is important to ensure that the message constitutes at least 50% of the total message package size. For this reason, it is important to reduce the size of package control and management compartments in subsequent message packages, and to use a random number update size of about ⅛ of the message size. The lower the random number information size, the higher the rate of message transfer and the lower the payload overhead. For increased security, larger update compartments may be used. Advantageously, the initial synchronization exchanges between machines may usually entail small messages, therefore allowing for large random update compartments and an acceleration of the RNG update process. It may also be possible that for any given or randomly predetermined message/random compartment length, the proportion of message to random compartment sizes may be determined randomly. For example in specifying that random updates should be 10% of message size, if given a total size of 101 bytes, a random number between 0 and 255 (modulus 21) may determine the random sequence size, and the residue may be filled in by next set of values from the message. Alternatively, the calculation could be applied to the message. Should the message be unable to complete the required compartment, the shortfall may be covered by the random updates. The finalized assignment may then be recorded in the package control section. In a similar manner, the values contained in the management section, may also be determined randomly. In addition, the management section may be removed completely, and values within the random update sequences may be used. As mentioned above, update sequences may be subject to independent encryption prior to being inserted into the message package. In addition, the message section may contain a multitude of messages and other encrypted sequences, which may be multiplexed or randomly configured prior to insertion into the message section. Thus the present invention may be used to increase and protect the security characteristics of existing cryptosystems.
Message control information may also include error correction codes, message compression configurations, or message source identifiers or demultiplexing instructions.
The present invention operates through the use of cryptographic or encryption cycles. Encryption cycles may include the creation, composition, encryption and decryption of one or more message packages, and may include the post transmission activities of message packages such as the updating of RNGs and cryptosystem management information in preparation for the next encryption cycle. Encryption cycles are almost always dependent on previous encryption cycles in that the specific cryptographic activities for a specific cycle are determined in previous cycles, and need not be the immediately preceding cycle, since some cycles may not contain message control or cryptosystem management information, and may be composed of message information only. Once a message package has been composed, it may be subjected to an encryption cycle, which includes the application of a predetermined sequence of one or multiple encryption or super-encryption mechanisms, which may include XORing or performing a mathematical operation on the contents of the package using one keystream, shuffling the contents with the same or a different keystream, and may include further, additional or repeated encryption operations using one or more keystreams. Upon receipt at the receiving end, operations may be applied in reverse order to disassemble the package, assign it to the appropriate message streams or destinations. Following that, both stations may update RNGs appropriately, process changed RNG values to upper level RNGs, and change predetermined characteristics of the encryption system in preparation of the next encryption cycle.
Central AuthorityCertain embodiments of the current invention include, but are not limited to embodiments where two or more independent entities in a network have pre-established secured communication links with a common entity or group of entities whose function may be to serve as a central authority for the identification, authentication and verification of network entities. The central authority may serve as a central communication destination where one entity may search for another entity and request a communication link to that entity via the central authority. The central authority may then, using dedicated or non-dedicated communication links to both entities, verify the identities of both parties and serve as a broker by receiving an initial communication or communication initialisation sequence such as that previously mentioned, from one entity and transferring it to the other, or it may send identical communication sequences to both parties. Once connected, the parties may or may not communicate independently of the central authority. The central authority may further be used in the event of a communication failure, or to verify additional information, or to temporarily keep communication records sent by one entity to another where one of the entities is not currently connected.
Certain embodiments of the current invention include, but are not limited to embodiments where a constant secure communication link may be maintained between a secured network entity and a central authority by way of an independent management process at a network node. The operational characteristics of this management process may be subjected to the control of the central authority and may be independent from direct entity (user, machine, operating system etc.) interaction. This communication link may also be used as an additional backup facility, or used to transfer central authority authentication or identification information when required.
Description of the Preferred EmbodimentMany different embodiments and variations of the various components of the invention have already been mentioned above. Certain embodiments of the current invention include, but are not limited to a preferred embodiment which incorporates the various methods, systems and components described above in a preferred manner due to efficiency, effectiveness, advantage, risk or perfect secrecy considerations.
With
Primary Generators 607 609—Each station will have one primary Hi-Generator which will serve random sequences to secondary generators. For scalability purposes, more primary generators will be added to serve blocks of secondary generators. Primary generators will have a single array of random size between 1024 and 2048 bytes and 2 index pointers of 2 bytes each. Each primary generator will intermittently receive small entropy inputs from true random sources such as timings between user key presses or the lower bits of network package arrivals. Primary generators will be instantiated with a saved encrypted RNG-state file which is decrypted with the user's password.
Communication Link 603—The communication link shown in
Initialisation Sequence 615—The station initiating the first communication link will use its Primary RNG to create a random number sequence to determine the constitution of the secondary RNG, which is usually an array of 2048 bytes, give or take 512 bytes. Secondary arrays are also populated using the same number sequence. The key component is provided using a symmetric key of any size that has been provided for the communication, or a Diffie-Hellman-Merkle key exchange using a multiplicative group of integers modulo p where p is not necessarily prime. The public key will be used by the receiver to return two sets of entropy update sequences. The first sequence will be applied to the secondary RNGs whilst the second will be used to select valid and invalid sets of values during decryption/encryption by using the separate digits in numbers (e.g. 4231, may mean take 4, skip2, take 3, skip 1, or alternatively, take 4, skip 6, take 2, skip 8 etc.). The initial configuration, the RNG data and the key will all be transmitted by the sender. With a central authority, this information can be distributed using the Central Authority as a broker. Once the first channel has been synchronized, having gone through a synchronization phase of about 8192 bytes, the receiver will repeat the activity, and send the sender an initialisation package for the second channel. Once the second channel is completed, the sender will reinitialise the first channel again. With a Central authority, all information is transferred between the two stations by a broker, so initialisation routines may not require large initial synchronization exchanges.
Secondary RNGs 608 and 610—Secondary Hi-Generators will have internal arrays of at least 2048 bytes. Arrays will be able to change array sizes between encryption and decryption cycles depending on the random entropy updates they receive from primary RNGs or if receiving messages, from the other linked and synched secondary RNG. RNGs will apply their updates after every cycle, and updates will change the values in the 3 index pointers. For scalability, and to ensure there is no contention, separate sections of the array may be interacted with by multiple processes. In addition, once updates have been applied to a section of the secondary RNGs internal state, the replaced values will be sent to the station's primary RNG. Secondary generators are instantiated with every communication link. When links are terminated, the generator is closed down, but not before its state is recorded and saved locally, and encrypted by another secondary generator. This allows connections to be stood up immediately, and serves as a backup in case of errors.
Data Multiplexer 605 and 612—Responsible for packaging and unpackaging transmission information in units of bytes. A random number sequence from the primary RNG will determine the overall package size, and the data/entropy balance which is usually 8-to-1 or 12% entropy for 100% data. Data Multiplexers at the sender will get the appropriate sized data from the Data Source 604 and the entropy from the Primary generator. Data Multiplexers process packages in terms of their size. Packaged data has no impact on the overall encryption process, except to occupy data blocks in the message package. Multiplexers may also perform additional activities such as data compression and error control coding. After sending the packages to be encrypted, Multiplexers will send the entropy updates to the Secondary RNG for application. Data Demultiplexers will receive decrypted packages and pass updates to secondary RNGs, and data outputs to the Data Destination buffers.
Encryptors/Decryptors 606 and 613—Encryptors receive message packages from Data Multiplexers and apply a super-encryption process using 2 separate keys, one for an XOR and the other to shuffle the resulting cyphertext. Two random bytes will determine the starting point for linked values, and the increment to be applied. Since the message package includes a message type ID, this may be used to vary the encryption sequence, and allow for additional encryption algorithms to be introduced. In order to expedite encryption, both XOR and shuffle operations may be performed off two output bytes from the secondary RNG. Decryptors perform the identified XOR/shuffle sequence in reverse order.
The encryption operations would essentially be operating on a variable or fixed length package. Each package may contain datagram elements for internal routing, control sections for functions such as MAC authentication or error correction, a message space, and a random number generation update. Control functions could be message size. All messages would be as large as the largest datagram, with free space padded out by random number generation updates. These updates considerably “pollute” any cryptanalytic attack. One issue to consider is that states as large as 1024 bytes can be used. Under ideal conditions, it would be best to ensure that the package is at least 700 bytes or 700 bits long, preferably 1024 or 1536 bytes. The logarithmic shuffling encryption operation has the unusual characteristic that beyond 700 bytes, there are more possible shuffle operations possible than the total variation of possible message contents. Issues to consider are that decryption operations must be performed in reverse, so the state overlap issue must be kept to a minimum. It is possible to allow for single state manipulation, but then large sets of values must be held in memory and applied in reverse, defeating the secret state directive. Keeping them separated avoids collisions. To protect against external manipulation or changes, a verification value or hash can be applied to the current datagram, but that adds additional overhead. Ideally we are looking to keep the random generation updates as small as possible and apply them to random index points of the PRNG state. To save space, the first random numbers or a predetermined sequence determined by previous control transmissions can be used to determine randomization for the next sequence. The instructions for updating the sequence are sent in previous packages. Whilst seeds seem to be the norm, this essentially just restarts the PRNGs and reduces the overall entropy of the solution. Every PRNG update serves as a pollutant, as instructions for the next update and as the content of the state change. Once transmitted, updates will be applied at the remote secondary RNG which has been linked to the communication link. Naturally, in order to promote recycling of values, remote primary PRNGs will have portions of their state values changed during the feedback loop at the remote site. To prevent this, each encryption engine will operate an independent secondary PRNG which will serve cryptographic administration and maintenance functions. One of the maintenance functions is to temporarily encrypt all backup link functions and amend any feedback loops if required through independent internal XOR operations.
Claims
1. A method of providing a secure communication link between system nodes, using a cryptosystem, the method comprising:
- providing a sequence of identical initial seed key values to sending and receiving system nodes,
- providing a sequence of message data values at a said sending system node to be communicated to any said receiving system nodes,
- generating a first sequence of pseudo-random or random entropy values at a sending system node,
- joining, combining or concatenating any said random entropy values to any said message data values in an encryption block having random or predetermined characteristics,
- generating a second sequence of pseudo-random or random key values at said sending system node and receiving system node, based on said seed key values,
- encrypting the contents of said encryption block in accordance with said second sequence at said sending system node and transmitting the encrypted values of said encryption block to said receiving system node,
- decrypting the received encrypted contents of the encryption block at any or all receiving system nodes in accordance with said second sequence, thereby extracting the message data values and the first sequence of random entropy values,
- augmenting, updating, or increasing the uncertainty or entropy of said cryptosystem in a symmetric or asymmetric manner at both the sending and receiving system node(s) using said first sequence of random entropy values.
- repeating if required, the steps above, for any subsequent message, entropy, control or key data which may need to be transmitted securely.
2. The method as set forth in claim 1 further including the step where the generation of the first sequence of random entropy values and second sequence of random key values is performed using one or more true or pseudo-random number generators, each having one or more internal states, variables or output sequences.
3. The method as set forth in claim 1, further including the case where the sequence of initial seed key values is not used as a seed key to seed a random number generator and produce a sequence of random encryption key values, but is used directly as the encryption key for any required encryption and decryption operations with regards to the encryption block by the respective system nodes, and where the first sequence of random entropy values is used to generate a new sequence of seed key or encryption/decryption key values as a function of the first sequence of entropy values and the previous sequence of seed key values (or encryption/decryption key values) by performing a logical operation on the previous seed key or encryption key using the first sequence of random entropy values.
4. The method of claim 1 wherein the steps relating to the augmentation, updating or increasing of the uncertainty of said cryptosystem further includes but is not limited to determining, altering, encrypting or performing a logical operation on any of the following:
- any internal variables of the cryptosystem;
- the selection and number of any encryption ciphers or algorithms;
- the sequence of encryption ciphers or algorithms;
- the initial seed key values;
- the internal state(s) of any pseudo-random or random number generator(s);
- the characteristics or composition or values of any said pseudo-random or random number generator(s) or their randomly generated sequences;
- the composition and characteristics of any encryption blocks;
- or any other component or cryptographic primitive which may be used in the cryptosystem.
5. The method of claim 1 wherein the step relating to the augmentation, alteration, determination or encryption of the composition, internal state(s), initial seed values, variables or outputs of any used random number generators using an additional random entropy value sequence other than any initial seed key values in a continuous manner, may occur independently in its own right as a separate invention relating to a new means of generating random numbers.
6. The method of claim 5 wherein the random number generators may be used independently from any encryption scheme but may be used purely as a means of probabilistic random number generation, and where they are composed of one or more internal states or variables of predetermined or random length or composition, such states being populated with random values, and where the values contained within the current internal states may be periodically amended, altered or encrypted in a random or predetermined manner using an independent or alternative source of entropy. In addition, the length of the internal states may be altered periodically by adding or removing values.
7. The method of claim 4, wherein the augmentation, alteration, determination or encryption of the internal state(s) or variables of any random number generator is accomplished by selecting specific values of the internal state(s) or variables using values in the first sequence of entropy values, and by performing a logical XOR or modular addition operation on said internal state or variable values, using the first sequence of entropy values.
8. The method of claim 5, wherein the said entropy update, augmentation or increase of any random number generators may result in an output set of values which may then be used to further update, augment or increase the entropy of any other random number generator used by the cryptosystem.
9. The method of claim 1, wherein the generation of any first or second random value sequences, may include the generation of a multitude of random sequences.
10. The method of claim 1, wherein said joining or concatenation of said message, entropy, control and other values may include additional symmetric or asymmetric key values and entail the joining or concatenation of various sections or lengths of said values in a predetermined or randomly determined manner, and may include the joinder of one, part or a multitude of messages, entropy sequences or other control or key values from a multitude of sources.
11. The method of claim 10, wherein the individual message, entropy, control or key values may be separately or jointly encrypted prior to being joined, using the initial seed key values or part thereof.
12. The method of claim 1, wherein the generated first sequence of random entropy values and second sequence of random key values may be constituted of a multitude of entropy or key value sequences from one or more said random number generators.
13. The method of claim 1, wherein the step relating to the encryption of the contents of the encryption block, may be accomplished using multiple encryption steps, algorithms or ciphers, the composition and sequence of which may be predetermined or may be randomly determined at runtime by values contained within the entropy sequence, or by values contained within the second generated sequence of random key values.
14. The method of claim 13 wherein the multiple encryption and decryption steps, algorithms or ciphers may be composed of a sequence of substitutions, XORs, transpositions, other cryptographic primitives or by performing a logical operation on the encryption block using the second sequence of key values.
15. The method of claim 1, wherein the step relating to the decryption of the encrypted contents of the encryption block, may be accomplished by repeating the encryption sequence of said steps, algorithms ciphers or logical operations in reverse.
16. The method of claim 1 wherein control values may be predetermined or randomly determined using values from the first sequence of random numbers or any other sequence, and wherein control values may be used to determine the encryption block characteristics, including encryption block length, message length, random entropy length, block length, and key length in the current or future encryption blocks.
17. The method of claim 16 wherein the encryption block or control values may contain error correction or message authentication verification information, or where the encryption block may be used to transfer user authentication information, user mutual authentication information, or additional symmetric or asymmetric keys which may be used for such purposes, or where the cryptosystem may be used by a Central authority as a means of brokering a connection between any system nodes which it is already in communication with.
18. The method of claim 16 wherein the composition of the cryptosystem PRNGs may be determined by the initial sequence of seed key values and where any internal states or variables may be populated using said seed key values. Alternatively, or in addition, any required RNGs may be initialised using an initial random entropy sequence which may or may not be encrypted with an asymmetric key pair, or where the random entropy sequence may be used to specifically select various outputs values from a designated PRNG as a means of populating the internal variables or internal states of any other required PRNGs (such as when an additional communication channel may be required), or in order to determine their composition.
Type: Application
Filed: Jul 29, 2016
Publication Date: Feb 2, 2017
Patent Grant number: 10348704
Inventor: Helder Silvestre Paiva Figueira (London)
Application Number: 15/224,492
